The GNOME project at 15

When you create a consumer operating system, you have to make some compromises. You have to have good support for proprietary applications as well as open source ones. Most consumers aren't interested in being part of a community. They're just interested in having stuff work without too much effort. So you make it difficult to modify the base system. No, this doesn't mean you have to totally lock down the system the way some vendors do, but at the very least, you have to make it difficult for Grandma to go off the rails. You have to have good hardware support, so you work together with OEMs. Some of them might want to pre-install applications on the system... so you let them.

People keep treating this like it's a technical problem, but it's really not. From a technical perspective, Android has everything you could want. Apps are (mostly) written in a common, safe, high-level language-- Java. Both open source and commercial third-party apps are supported pretty well. There's a touchscreen GUI that people are pretty happy with. Android has a capability-based security model. All the stuff that the GNOME guys are talking about doing for "GNOME OS" is stuff that Android has already done.

Capitalism tends to increase our isolation from the things we consume. Instead of raising our own livestock, slaughtering them, and cooking them ourselves, we go to McDonald's and buy a pre-wrapped, pre-cooked meal in a box. Instead of assembling our own PCs out of parts and hacking together a kernel and applications for them, we pick up a phone with the whole operating system pre-installed, plus a few apps.

If GNOME was successful, would it really be any different than Android is now? It's better for open source developers to push the frontiers in other ways than to chase Google or Apple's taillights.